Example · Muon g−2 2021
A 4.2σ discrepancy — with an important boundary
The familiar number is meaningful only relative to the theoretical benchmark against which it was calculated. That boundary is part of the result.
In 2021 a 4.2σ tension was reported between experiment and a particular Standard-Model benchmark. The test finds that this precisely benchmark-relative statement is supported. It does not establish “new physics” or adjudicate every other theory calculation.
The six questions
What claim was tested?
The original wording is preserved for auditability.
Muon g−2 Collaboration (2021): 4.2σ experiment–theory tension relative to the declared data-driven HVP / WP20 benchmark synthesized by Aoyama et al. (2020).
What evidence state applied?
7 April 2021. The comparison is fixed to the declared data-driven HVP / WP20 reference. The contemporaneously relevant BMW lattice calculation remains visible as important context, but it is not silently substituted for the benchmark of the tested claim.
What was the result?
PASS. No applicability-relevant deficit remained for the exact benchmark-relative 4.2σ statement.
Why?
The numerical statement was explicitly tied to one theoretical benchmark, and that qualification was preserved. The result can therefore close without making the statement stronger than its evidential basis permits.
What does this not mean?
PASS does not mean that the Standard Model is confirmed or refuted, that all theoretical determinations agree, or that new physics has been established. Only the declared benchmark-relative 4.2σ statement passed.
What would trigger a new test?
A new theoretical benchmark, a changed reference calculation, materially new evidence, or a differently formulated claim. The 7 April 2021 result remains as its own historical record.
Show technical details
Run ID: RMuon
P2 route: L3→L2 benchmark-relative attribution/comparison
Comparator: data-driven HVP / WP20; BMW retained as scope context.
Evidence date: 7 April 2021